Diesel vehicles may be equipped with an exhaust gas treatment system which may include, for example, a urea based selective catalytic reduction (SCR) system and one or more exhaust gas sensors such as nitrogen oxide (NOx) sensors, at least one of which may be disposed downstream of the SCR system. When the SCR system becomes loaded with urea to a point of saturation, which varies with temperature, the SCR system may begin to slip ammonia (NH3). The NH3 slip from the SCR system may be detected by the tailpipe NOx sensor as NOx resulting in an inaccurate NOx output which is too high. As such, the efficiency of the SCR system may actually be higher than the efficiency determined based on the inaccurate NOx output.
US 2012/0085083 describes a method for estimating NOx conversion using a polynomial model that also allows for the NH3 concentration at the downstream tailpipe NOx sensor to be estimated. As described therein, temporal sensor signatures of a feedgas NOx sensor located upstream of the SCR and a tailpipe NOx sensor located downstream of the SCR are quantified and fit using a polynomial model that enables estimation of NH3 slip and NOx conversion efficiency. However, because the method uses a segment of each sensor signal for processing, a time lag exists between the acquisition of each NOx sensor output signal and the allocation of downstream NOx sensor output to NOx and NH3. When the time lag is combined with the polynomial fitting algorithm described, which may be prone to localized estimation errors, realization of a real-time NH3 slip detection system by the methods described would be difficult to implement.
The inventors have recognized disadvantages with the approach above and herein disclose methods for the real-time control of ammonia slip in an engine exhaust system. Methods described use transient responses of a NOx sensor to identify the rates of change of a NOx signal. Then, a processor further uses the rates of change of the NOx signals to determine how the downstream tailpipe NOx sensor is expected to change based on the exhaust flow upstream of the SCR, which allows allocation of the tailpipe NOx sensor in the manner herein described. The inventors have further recognized that a tradeoff exists between the sensitivity and accuracy when processing transient signals. For this reason, the disclosed segment length method increases signal sensitivity while maintaining a high degree of accuracy, particularly at low exhaust temperatures, which allows for enhanced signal processing during lightly driven engine cycles. In addition, the inventors have further recognized that the envelope technique may have a reduced sensitivity than the segment length method in some instances since phasing errors cause the envelopes to be larger due to the phasing. Moreover, the envelope method is susceptible to being disabled during directional changes of the transient signal, which may result in a reduced accuracy in some cases. Thus, the segment length method disclosed is more robust to noise factors and phasing errors, and may thus have a higher sensitivity without a substantial reduction in accuracy. As one example, the inventors found that the sensitivity of the segment length method to ammonia detection is ˜10 ppm whereas the sensitivity of the envelope method is ˜25-30 ppm. In one particular example, the exhaust system includes two NOx sensors that continuously monitor the exhaust gas flow upstream and downstream of an SCR device. Then, when entry conditions of the engine system are met, for example when the SCR device is above a temperature threshold, the rate of change of the upstream feedgas NOx sensor is combined with a current tailpipe reading to estimate the rate of change of the tailpipe NOx sensor expected based on the feedgas signal slope. The expected tailpipe NOx signal is then compared with the actual NOx signal in order to allocate the NOx sensor output to NOx and NH3 using one or more of the envelope method and segment length method disclosed herein and described in detail below.
In another example, a method is provided that comprises allocating a NOx sensor output to each of NH3 and NOx based on an upstream NOx rate of change and a downstream NOx rate of change relative to the SCR emission device, which thereby allows the amount of reductant delivered to the engine exhaust to be adjusted based on the relative sensor signals. Because the method uses transient responses of the upstream and downstream NOx sensors, in addition to the expected NOx signal, it is therefore possible to achieve a high level of NH3 detection. In this way, it is possible to provide enhanced allocation of the NOx sensor output in order to determine the relative NOx and NH3 levels in the exhaust system.
The present description may provide several advantages. In particular, the segment length approach may allow for the real-time detection of NH3 slip with a high level of detection sensitivity without high feedgas NOx interventions while preserving detection accuracy. Thus, NH3 slip can be detected while a vehicle is in operation and corrective measures taken based on the current state of the exhaust system. Furthermore, because the detection sensitivity is increased, high levels of NOx are not required in order to determine the allocation of the tailpipe NOx sensor output to NOx and NH3.
The above advantages and other advantages, and features of the present description will be readily apparent from the following Detailed Description when taken alone or in connection with the accompanying drawings. It should be understood that the summary above is provided to introduce in simplified form a selection of concepts that are further described in the detailed description. It is not meant to identify key or essential features of the claimed subject matter, the scope of which is defined uniquely by the claims that follow the detailed description. Furthermore, the claimed subject matter is not limited to implementations that solve any disadvantages noted above or in any part of this disclosure.